Small Grants Spotlight: Loreto Cox & Natalia Garbiras-Díaz

Loreto Cox
Natalia Garbiras-Díaz

This figure shows the most common terms used by treated respondents in their responses to our open-ended priming question.
This figure presents the results of the estimation of the marginal means for the attribute levels of candidate slogans (i.e., anti-establishment appeal or neutral).
This figure shows the results of estimating the marginal means for combinations of anti-establishment rhetoric and anti-democratic phrases by treatment condition and sample. “Neutral” corresponds to candidates who do not use anti-establishment slogans, while “Anti-establishment rhetoric” corresponds to those who do. “Anti-democratic phrase” corresponds to candidates who adopt an anti-democratic stance. We separate the analysis between subjects assigned to the treatment (Figure a) in the main experiment and those assigned to the control (Figure b). We test whether the difference between the estimated marginal means in each country is statistically significant. For the treatment group: Colombia (β̂diff =0.05, p-value<0.01), Peru (β̂diff =0.03, p-value<0.1), Pooled (β̂diff =0.04, p-value<0.01); for the control group, the differences are not statistically significant.